In this project, I developed an image classification model to identify plant species using a Telegram bot interface as my final Master Thesis project. Users can send images of plants to the bot, which then classifies the species and provides relevant information. To enhance the user experience, the bot also sends reminders to irrigate the plants based on their specific watering needs.
The project incorporates robust MLOps features, leveraging the DVC (Data Version Control) library to manage datasets, model versions, and experiments efficiently. This ensures reproducibility and scalability, allowing for seamless updates and improvements to the model. Additionally, automated documentation is maintained using Sphinx, providing clear and comprehensive project documentation that is continuously updated as the project evolves.
Overall, this project not only demonstrates advanced machine learning and image classification techniques but also showcases effective MLOps practices and user-centric design through the integration of automated reminders and thorough documentation.